lgy@phys.washington.edu (Laurence G. Yaffe) (08/06/90)
On July 14, Laurent Siebenmann <SIEBENMANN@LALCLS.DECNET.CERN.CH> wrote:
-
- This problem occurs naturally with mathematical
- displays that accidentally exceed line length. The typist
- casually throws in \\ when the overrun occurs, and
- disaster ensues if the elastic (self-sizing) \left( and
- \right) are separated by the \\. To avoid a TeX error,
- you would have to balance the "\left(" with an invisible
- "\right." on the same line and similarly for "\right)".
- But that may lead to parentheses of different sizes, a
- minor tyographic disaster rather than a TeX error.
-
- A nasty problem that typist repeatedly lose time
- on. Indeed many math typists completely abandon use of the
- self-sizing mechanism to avoid this one hitch.
-
I stopped using \left and \right long ago due to both the above-mentioned
problem plus the perception that 90% of the time the delimiters produced by
\left & \right are too large. This is a subjective judgement, but I feel that
\left( & \right) very frequently produce parentheses which are so large that
they, completely undeservedly, become the primary focus for one's visual
attention. Specifically, I almost always use \Big delimiters when \left
would have chosen the next larger size. I feel that that delimiters which
are slightly smaller in size than the tallest item they enclose are more
than sufficient to adequately set-off the enclosed expression.
(Unfortunately, all technical journals which use TeX (or troff) for their
typesetting that I've seen accept the default behavior of \left & \right.
In my opinion, this problem is the largest single cause of the degradation
in viaual appearance of these journals produced by the abandonment of
hand-set type.)
--
--------------------------------------------------------------------------
Laurence G. Yaffe Internet: lgy@newton.phys.washington.edu
University of Washington Bitnet: yaffe@uwaphast.bitnet